Robotic hand self-righting with visual feedback

I gave GPT-6 Astra (Ultra) access to a Wuji2 hand and a third-person camera, and asked it to get the hand upright. Watching it use its fingers to lift itself off the table felt a little like watching a baby learn to stand. The hand stayed upright after the motors were disabled.

Ultra succeeded in about 3 minutes. An earlier run using Max Effort failed after about 8 minutes of repeated attempts and adjustments.

One prompt

We have connected a third-person camera and a Wuji2 robotic hand for control. Please control the robotic hand to move itself into an upright configuration, with its base positioned at the bottom, rather than lying flat as it is now. You need to figure out on your own how to adjust the robotic hand pose to achieve this.

Please record a video while operating.

How Astra produced the motion

Astra used Wuji's existing MJCF model in MuJoCo, added a floating base and tabletop contact, and searched over sequences of joint poses. I double-checked the control setup: it did not build a conventional inverse-kinematics pipeline.

On the real hand, those poses were sent as joint-position targets through Wuji's SDK. Camera images, joint readings, and IMU feedback were used to check the motion and adjust between attempts. The fingers and thumb provided support as the hand lifted and transferred its weight onto the base. Coordinating those contacts was central to getting it upright.

After the motors switch off

The final pose appears to use its support contacts to remain upright, with joint friction potentially contributing. Reaching this pose and maintaining it after actuation ends are distinct parts of the task.

Interpreting the result

I distinguish the agent's search over poses from the controller that executes them. The observed success belongs to this combined system: the supplied hand model, contact simulation, feedback, and joint controller all contribute. This demonstration does not isolate which component was necessary, or establish that Astra inferred the hand's dynamics accurately.

The Ultra and Max runs also differ in their search histories, so their completion times do not constitute a controlled comparison of reasoning-effort settings.

Contact modes and terminal stability

The upright, unpowered pose suggests an objective beyond reaching a joint configuration: reaching a mechanically stable support state. Future search methods could plan the contact modes that lead to that state, specifying which fingers provide support and when those contacts change. Feedback could then revise the transitions when tabletop interaction differs from simulation.

My posts on X: demonstration · simulation and control details.

@misc{dou2026handselfrighting,
  author = {Dou, Zhiyang},
  title = {Robotic hand self-righting with visual feedback},
  year = {2026},
  month = sep,
  howpublished = {Research Notes},
  url = {https://frank-zy-dou.github.io/blog/wuji2-hand-stands-up/}
}